Apoptosis is a genetically-encoded form of cell death that is required for tissue homeostasis. Consequently, mutations that suppress apoptosis can promote tumor progression. Because many anticancer agents induce apoptosis, mutations in apoptotic pathways can also produce a pleiotropic resistance to cancer therapy. Previous studies have shown that oncognic changes, as mimicked by the adenovirus E1A oncogene, can sensitize otherwise resistant cells to the induction of apoptosis by diverse agents. Other mutations--for example, inactivation of the p53 tumor suppressor gene--can suppress apoptosis. Together, our studies have identified p53 as an important regulator of apoptosis in tumor cells and demonstrated that p53 mutations can promote oncogenic transformation, tumor development, and drug resistance by suppressing apoptosis. Despite their clinical importance, the molecular mechanisms of apoptosis in tumor cells remain poorly understood. Loke cells expressing E1A, tumor cells are typically more susceptible to apoptosis than normal tissues. Genes such as p53 and bcl-2 appear to regulate apoptosis in tumor cells, but their mechanism of action is unknown. A novel class of cysteine proteases related to interleukin-1 beta-converting enzyme may be essential components of the apoptotic machinery, but their role in cancer and chemosensitivity is undefined. The objective of this project is to further elucidate the molecular control of apoptosis in tumor cells. Our strategy is to apply principles from classical genetics to systems amenable to mammalian cell biology. By combining null mutant cells with powerful gene transfer methods, we plan to genetically dissect the apoptotic pathways active in E1A- expressing cells and determine how genes that regulate apoptosis interface with essential components of the apoptotic machinery. Ultimately, thee studies will identify molecular determinants of radiation and chemosensitivity and perhaps novel targets for therapeutic intervention. These studies are compatible with the overall goals of the program project.